The example from
http://gking.harvard.edu/zelig/docs/Example8.html
generates an error message at the first call to plot:
plot(s.out)
Error in plot.ci(x, xlab = "", ...) :
plot.ci() is valid only for non-categorical, univariate response models.
After reading the help files I sort of would expect this error, but since it's in both the ologit and oprobit example help files, I'm not sure I'm correct. Any suggestions?
I'm running the latest version in R 2.3.1.
Thanks.
Regina Rendas-Baum
-
Zelig Mailing List, served by Harvard-MIT Data Center
Send messages: zelig(a)lists.gking.harvard.edu
[un]subscribe Options: http://lists.gking.harvard.edu/?info=zelig
Zelig program information: http://gking.harvard.edu/zelig/
The example from
http://gking.harvard.edu/zelig/docs/Example8.html
generates an error message at the first call to plot:
plot(s.out)
Error in plot.ci(x, xlab = "", ...) :
plot.ci() is valid only for non-categorical, univariate response models.
After reading the help files I sort of would expect this error, but since it's in both the ologit and oprobit example help files, I'm not sure I'm correct. Any suggestions?
I'm running the latest version in R 2.3.1.
Thanks.
Regina Rendas-Baum
-
Zelig Mailing List, served by Harvard-MIT Data Center
Send messages: zelig(a)lists.gking.harvard.edu
[un]subscribe Options: http://lists.gking.harvard.edu/?info=zelig
Zelig program information: http://gking.harvard.edu/zelig/
When I attempt to supply a long variable list as part of the formula
in zelig(), I am getting an error:
> z.mod92.a <- zelig(as.factor(presvote) ~ whitesouth + relft + soph2 + pid + age + unionhh + black + married + female, model='mlogit', data=smallframe)
Error in parse(file, n, text, prompt) : syntax error in "cbind(
id(presvote, "2") ~ whitesouth + relft + soph2 + pid + age + ,"
Removing one variable, however, solves the problem:
> z.mod92.a <- zelig(as.factor(presvote) ~ whitesouth + relft + soph2 + pid + age + unionhh + black + married, model='mlogit', data=smallframe)
Changing the variables involved seems to make no difference; it is
apparently just the length of the variable list that matters.
Incidentally, I can run the model fine using vglm (through VGAM)
directly.
traceback() says:
12: parse(text = x)
11: eval(parse(text = x)[[1]])
10: formula(eval(parse(text = x)[[1]]))
9: switch(mode(x), "NULL" = structure(NULL, class = "formula"),
character = formula(eval(parse(text = x)[[1]])), call = eval(x),
stop("invalid formula"))
8: formula.default(object, env = baseenv())
7: formula(object, env = baseenv())
6: as.formula(paste(lhs, rhs))
5: cmvglm(formula, mf$model, mf$ones, ndim, data, fact)
4: zelig2mlogit(as.factor(presvote) ~ whitesouth + relft + soph2 +
pid + age + unionhh + black + married + female,
"mlogit", list(weight = c(1.7249, 0.4922, 1.1457, ....)), 1)
3: do.call(fn1, list(formula, model, dat, N, ...))
2: zelig.default(as.factor(presvote) ~ whitesouth + relft + soph2 +
pid + age + unionhh + black + married + female,
model = "mlogit", data = smallframe)
1: zelig(as.factor(presvote) ~ whitesouth + relft + soph2 + pid +
age + unionhh + black + married + female,
model = "mlogit", data = smallframe)
Thanks in advance,
Chris
--
Christopher N. Lawrence <c.n.lawrence(a)gmail.com>
-
Zelig Mailing List, served by Harvard-MIT Data Center
Send messages: zelig(a)latte.harvard.edu
[un]subscribe Options: http://lists.hmdc.harvard.edu/?info=zelig
Zelig program information: http://gking.harvard.edu/zelig/